Andy Hon Wai Chun, City University of Hong Kong; Steve Ho Chuen Chan, Garbbie Pui Shan Lam, Francis Ming Fai Tsang, Jean Wong, and Dennis Wai Ming Yeung, Advanced Object Technologies Limited
This paper describes the Rostering Engine (RE) that we have developed for the Hospital Authority (HA), Hong Kong as part of their Staff Rostering System (SRS) using AI constraint-programming techniques. The Hospital Authority manages over 40 public hospitals in Hong Kong. With close to 1500 wards total, the amount of resources needed to produce weekly staff rosters for each ward is tremendous and extremely time consuming. Previously, most staff rosters were generated manually. Without computer records, it was difficult for HA to produce workforce statistics or to improve resource efficiency. In early 1997, the Hospital Authority embarked on a strategic Staff Rostering System project to provide automation support in managing their large workforce of over 48,000 full-time hospital staff. The Staff Rostering System performs ward-level rostering based on ward-specific constraints, staff requests, and work patterns. Version 1 of the system was completed early 1998 and Version 2 was released early 1999. The system is gradually being deployed in different public hospitals across Hong Kong.